On June 22, 2026, AzerTelecom and Telecom Armenia signed a bilateral agreement enabling mutual transit of international internet traffic through each other’s territory. The press release was four sentences long. The political weight of those four sentences — in a region where two wars have been fought in thirty years, where approximately 100,000 people were displaced in 2023, where prisoners remain in detention and a peace treaty remains unsigned — is considerably larger than the language used to describe them.
By Vladimir Tsakanyan, PhD · Center for Cyber Diplomacy and International Security · cybercenter.space
The telecommunications industry produces a specific category of announcement. Bilateral transit agreements, route diversity enhancements, and regional connectivity frameworks are announced with functional language stripped of drama and optimised for an audience of network engineers and infrastructure investors. The agreement signed today between AzerTelecom, Azerbaijan’s leading backbone internet provider, and Telecom Armenia, Armenia’s leading transit operator, conforms precisely to this template. The joint press release identifies the parties, describes the scope — mutual transit of international internet traffic for the purposes of route diversification and resilience in the South Caucasus — and closes with a commitment to regional digital infrastructure development. It is, by every conventional measure of the genre, an unremarkable document.
What makes it remarkable is what it does not say: that for the first time since the dissolution of the Soviet Union, international internet traffic from Armenia will be routed through Azerbaijani territory — across the infrastructure of a country with which Armenia has fought two wars in thirty years, whose military operation in September 2023 brought an end to Armenian presence in Nagorno-Karabakh and produced one of the largest and fastest civilian displacements in post-Soviet history, and with which a final peace treaty remains initialed but unsigned, legally incomplete, and domestically contested in both capitals.
The digital packet does not carry a flag. It carries the weight of everything that has happened between the two states since 1988 — and it is crossing the border anyway.
Thirty Years of Disconnection
To understand the agreement’s significance, it is necessary to understand the geographic and political architecture that shaped Armenia’s internet connectivity for the entirety of the post-Soviet period.
Armenia is a landlocked state bordered by four countries: Azerbaijan to the east and northeast, Turkey to the west, Iran to the south, and Georgia to the north. The border with Turkey has been closed since 1993, a consequence of Turkey’s political solidarity with Azerbaijan during the first Karabakh war. The border with Azerbaijan has been closed since the same period. The border with Iran has remained limited in economic and transit significance. The practical result is that Armenia’s international connectivity — in the physical sense of roads, railways, and cables — has been channelled almost entirely through Georgia, a single neighbour corridor whose geographic position as the only reliable land transit route for Armenian trade and communications has been both Armenia’s strategic vulnerability and Georgia’s consistent economic asset.
Internet connectivity has followed the same geographic logic. For thirty years, international internet traffic to and from Armenia has been routed through Georgian infrastructure, through Georgian transit providers, and out to the global internet through corridors that consistently bypassed the Azerbaijani territory that geographical proximity would otherwise suggest as a natural route. The political enmity between Armenia and Azerbaijan, formalised in the closure of the border and the absence of diplomatic relations, was written into the routing tables of the South Caucasus internet as comprehensively as it was written into the political agreements that governed state relations.
The agreement signed today does not merely add a new route. It erases, in one technical decision, thirty years of routing architecture that reflected the geopolitical conflict between the two states. The internet packet that crosses into Azerbaijan is going somewhere that the political relationship between the two states had, for three decades, made entirely inaccessible.
The Context of Conflict: What This Agreement Is Being Signed Across
The political weight of the AzerTelecom-Telecom Armenia agreement cannot be assessed without precise accounting of what has occurred between the two states in the period whose legacy this agreement is attempting, at the infrastructure level, to begin transcending.
The first Nagorno-Karabakh war, fought between 1988 and 1994, established Armenian control over the disputed territory and surrounding regions, producing approximately 600,000 Azerbaijani displaced persons and ending in a ceasefire that left the political status of the territory unresolved. The frozen conflict that followed lasted twenty-six years — punctuated by periodic skirmishes along the line of contact, diplomatic negotiations that produced no settlement, and the steady accumulation of grievance on both sides.
The second Karabakh war, in autumn 2020, lasted forty-four days and fundamentally altered the territorial balance. Azerbaijani forces, with Turkish military support and using Turkish-supplied drone technology whose operational effectiveness exceeded almost all prior assessments, retook significant portions of the territory Armenia had controlled since 1994. A Russian-brokered ceasefire on November 9, 2020 halted the fighting, established Russian peacekeeping forces in the remaining Armenian-controlled areas of Karabakh, and created a new territorial reality that neither side had fully internalised by the time the ceasefire was signed.
The September 2023 Azerbaijani military operation — a twenty-four-hour offensive that the Azerbaijani government described as anti-terrorist in character and that the Armenian government in Karabakh was unable to militarily resist — ended with the agreement for the dissolution of the self-proclaimed Republic of Artsakh and the departure of its Armenian population. Approximately 100,000 people — the near-entirety of the ethnic Armenian population of Nagorno-Karabakh — fled to Armenia in the days following the operation. The demographic transformation of the territory was, by any measure, total and swift.
Against this history — two wars, mass displacement, the end of Armenian political presence in a territory that carried profound historical and symbolic significance for Armenians worldwide — the routing of internet traffic through Azerbaijani infrastructure is not a neutral technical decision. It is an act of engagement with a state whose recent history with Armenia includes events whose full political and emotional weight has not been absorbed by either society.
Analyst note
The Armenian domestic political context in which this agreement is being received adds a dimension that the press release’s functional language cannot accommodate. Armenian opposition parties have characterised the normalization process with Azerbaijan — of which this telecom agreement is the most recent and most concrete step — as capitulation, as the formal acceptance of the territorial and political losses that the 2020 and 2023 military defeats produced. The opponents of normalization do not dispute the economic logic of the agreement. They dispute its political meaning: that by routing internet traffic through Azerbaijan, Armenia is normalising a relationship with a state that they hold responsible for the forced displacement of 100,000 Armenian civilians. The same agreement that appears, from the outside, as a pragmatic infrastructure decision reads, from the inside of Armenian domestic politics, as a symbol of what has been lost and what is being accepted.
The Asymmetry of Need and the Logic of Mutual Benefit
The political paradox of the Armenia-Azerbaijan telecom agreement is that both sides have genuine reasons to sign it, which is precisely what makes it possible at this moment and not at an earlier one.
For Armenia, the logic is structural vulnerability. A state dependent on a single transit corridor — Georgia — for its international connectivity is exposed to a category of geopolitical risk that any network engineer would immediately identify as unacceptable in any other context. Route diversity is a fundamental principle of resilient network architecture. Armenia’s thirty-year absence of route diversity through Azerbaijani territory was a political constraint that translated directly into infrastructure fragility: any disruption to Georgian transit routes — through political instability, natural disaster, or deliberate interference — would significantly impair Armenia’s international connectivity without the alternative routing that the agreement now provides.
For Azerbaijan, the logic is connectivity ambition. AzerTelecom has positioned itself as the South Caucasus region’s leading backbone provider and transit hub — a role it cannot fully realise without the route diversity that Armenian territory provides for traffic moving to and from specific regional destinations. The “Digital Silk Way” branding that accompanies AzerTelecom’s international positioning reflects an ambition to be the digital transit corridor connecting Europe and Asia through the South Caucasus — an ambition for which the normalization of routing relationships with Armenia’s operators provides genuine commercial value.
The mutual benefit structure is what creates the political space for the agreement to exist at all. Agreements between adversarial states are most achievable when both sides can identify concrete gains that are not zero-sum — where one side’s benefit does not come at the other’s expense. A route diversity agreement is almost a paradigm case of non-zero-sum infrastructure cooperation: Armenia’s connectivity improves, Azerbaijan’s transit capabilities expand, and neither benefit depends on the other’s loss.
The Normalization Sequence: This Agreement in Its Political Context
The Azernews analysis published simultaneously with the agreement’s announcement characterises the telecom deal as the latest in a sequence of normalisation steps that began with the August 2025 Washington peace summit where a peace agreement was initialed. The sequence deserves precise reconstruction, because this agreement’s existence is only comprehensible against its backdrop.
Following the Washington summit: Azerbaijan removed transit limitations for third countries to Armenia, reconnecting Armenia’s trade relationships with partners whose goods had previously been blocked from transit through Azerbaijani territory. Azerbaijan resumed oil supply transit through the rail link through Georgia — a commercial relationship that had been suspended during the period of hostility. Border delimitation discussions produced an actual delineation of portions of the border. An aide to the Azerbaijani president travelled to Armenia through the newly delineated border crossing — a gesture whose symbolic significance in a region where no senior Azerbaijani official had crossed into Armenia for decades was considerable despite its operational simplicity.
Each of these steps shares a structural characteristic with the telecom agreement: they are discrete, reversible, and low-visibility. None of them requires the formal signing of the peace treaty that both governments have been approaching for years but have not yet completed. Each of them creates a practical condition on the ground — a transit route reopened, a border crossed, a cable routed — that is marginally harder to undo than the step before it. The accumulation of these practical conditions is the normalisation process, conducted in the language of logistics rather than diplomacy, creating facts through infrastructure rather than through the treaty language that domestic political opposition in both countries would contest.
The most astute observation in the Azernews analysis is the one that identifies what is different about the telecom agreement from the preceding steps in the sequence: “Unlike other declarations, the routing of the internet is an either/or proposition.” The transit route that was opened can be closed again. The border crossing can be suspended. The oil supply can be redirected. But internet routing, once implemented in the configuration management of the networks involved, is a binary condition — the route exists in the routing table or it does not. The packet crosses the border or it goes another way. There is no intermediate position, no partial compliance, no face-saving ambiguity. The technical commitment is total and immediately verifiable.
Who Actually Signed: The Corporate Actor as Diplomatic Bridge
The most politically significant dimension of the AzerTelecom-Telecom Armenia agreement is identified in the Azernews analysis with unusual directness: “One of the most fascinating aspects of this agreement is who actually made it.” The agreement was not signed by the governments of Armenia and Azerbaijan. It was signed by two companies — telecommunications operators with commercial objectives, commercial relationships, and commercial accountability structures that are distinct from, though connected to, the political structures of their respective states.
This distinction matters more than it might initially appear. A government-to-government agreement in the current political environment — where the peace treaty remains initialed but unsigned, where Armenian opposition parties are challenging parliamentary election results, where prisoners from the 2020 conflict remain in Azerbaijani detention, and where the Armenian constitutional provisions that Baku regards as incompatible with peace remain in place — would carry a domestic political weight that could be difficult for either government to manage. A company-to-company commercial agreement, framed in the language of route diversity and network resilience, carries the same technical consequence with a significantly different political exposure profile.
The use of corporate actors to advance state objectives in the digital domain is a pattern documented extensively in the broader literature of technology governance: states whose political constraints prevent direct government-to-government action create the regulatory and commercial conditions in which their corporate entities can take steps that advance the state’s strategic interests without requiring the political accountability that direct government action would demand. Whether the AzerTelecom-Telecom Armenia agreement reflects a deliberate strategy of using corporate actors to advance a normalization agenda that governments cannot advance directly, or whether it reflects genuine corporate initiative operating in a political environment that has become sufficiently permissive to allow it, is a question whose answer may not be fully visible from the outside. The effect is the same in either case: infrastructure has crossed a border that politics has not yet formally opened.
The Unresolved Pain Points
The agreement’s existence does not resolve, and its language does not address, the conditions that make the South Caucasus political environment in which it is being signed one of the most complex in the post-Soviet space.
The fate of Armenian prisoners held in Azerbaijan — civilians and military personnel captured during the 2020 war and the 2023 operation — remains a major unresolved issue. International organisations including the International Committee of the Red Cross have repeatedly called for the return of all detainees. Armenia has made the release of prisoners a consistent demand in normalization negotiations. The absence of any movement on this issue, despite the normalization sequence that has produced the telecom agreement, is the single most concrete measure of the gap between the pace of infrastructure normalisation and the pace of humanitarian resolution.
The Armenian constitutional question — Baku’s demand that Armenia remove provisions of its constitution that it regards as implying territorial claims over Karabakh — has not been resolved and remains a formal precondition for the signing, as opposed to the initialing, of the peace treaty. The Armenian government’s position is that constitutional amendment requires a domestic political process that cannot be completed on a timeline determined by bilateral negotiations. The opposition’s position is that the demand itself is evidence of Azerbaijani maximalism. The gap between these positions is not bridged by any of the practical steps in the normalization sequence, including today’s telecom agreement.
The status of Karabakh itself — the territory, its infrastructure, its historical and cultural heritage, and the conditions under which its former Armenian inhabitants might or might not return — has been formally resolved as a matter of sovereignty (Azerbaijani) but not as a matter of the practical and humanitarian questions that the Armenian community and international observers have raised. The Armenian community of Karabakh, now displaced in Armenia, has not been part of any consultation about the normalization process that the agreement represents.
Analyst note
The telecom agreement’s framing — mutual route diversity and resilience enhancement — is technically accurate and commercially sound. It is also, in the Armenian domestic political context, a framing that elides the asymmetry of what the two sides are agreeing to in symbolic terms. Azerbaijan, by routing internet traffic through Armenia, is routing through the territory of a neighbour with whom it has signed a peace framework and whose political position has moved significantly toward accepting the post-2023 reality. Armenia, by routing internet traffic through Azerbaijan, is routing through the territory of a state it holds responsible for the displacement of 100,000 Armenians from a territory of profound historical and cultural significance. The commercial logic is identical for both. The symbolic content is not.
A Small but Precise Example of Cyber Diplomacy
The AzerTelecom-Telecom Armenia agreement deserves its place in any serious analysis of cyber diplomacy not because of its bandwidth or its technical specifications, but because it demonstrates, with unusual clarity, the mechanisms through which digital infrastructure can advance political relationships that formal diplomatic processes have not been able to move.
Cyber diplomacy, as a field, has developed primarily around its conflict dimensions: the attribution of state-sponsored attacks, the negotiation of norms for responsible state behaviour in cyberspace, the management of escalation in a domain where the boundaries between civilian and military targets are contested. These are real and consequential dimensions of the relationship between digital infrastructure and international politics.
The AzerTelecom-Telecom Armenia agreement illuminates a different dimension: digital infrastructure as a confidence-building instrument, a mechanism for creating facts on the ground that are verifiable, binary, and consequential in ways that declaratory diplomacy is not. The agreement does not resolve the disputes between Armenia and Azerbaijan. It does not address the prisoners, the constitutional question, the status of Karabakh’s former inhabitants, or the peace treaty’s completion. What it does is create a technical interdependency — a routing relationship — that makes the total disruption of connectivity between the two states marginally more costly than it was before the agreement was signed. Each packet that crosses the border is, in the most literal possible sense, a stake in the continuation of the practical relationship.
This is confidence-building at the infrastructure level. It is not the traditional confidence-building measure — the military observer, the hotline, the inspection protocol — but it shares the same operational logic: creating observable, verifiable, reversible-but-costly-to-reverse conditions that increase the practical cost of a return to hostility without requiring the political conditions that a formal treaty demands.
The agreement was announced in a four-sentence press release. Its significance is not in the sentences. It is in the routing tables.
Bottom Line Assessment
The AzerTelecom-Telecom Armenia internet transit agreement, signed on June 22, 2026, is simultaneously the most commercially unremarkable and the most politically significant telecommunications agreement concluded in the South Caucasus in thirty years. Its commercial logic — route diversity, resilience enhancement, regional connectivity development — is sound and mutual. Its political significance — the first routing of Armenian international internet traffic through Azerbaijani infrastructure since the Soviet dissolution, achieved by companies rather than governments, as part of a normalization sequence whose most concrete steps have been taken through logistics rather than diplomacy — is considerable.
It does not resolve the Armenian-Azerbaijani conflict. The prisoners, the constitutional question, the humanitarian status of Karabakh’s former Armenian population, and the unsigned peace treaty all remain as they were before the agreement was signed. The pain of what has occurred between the two nations — two wars, mass displacement, the end of an Armenian community’s presence in a territory they considered ancestral — is not addressed by a routing table entry, however symbolically significant that entry may be.
What the agreement does is demonstrate, at the smallest and most precise possible scale, the mechanism through which digital infrastructure can become a tool of political normalisation. Not through the grandiose language of peace, but through the functional language of network engineering. Not through the signing of treaties that domestic politics in both capitals cannot yet accommodate, but through the signing of commercial contracts that both sides can describe in terms of route diversity and resilience.
The packet crosses the border. The political relationship has not yet followed. But the infrastructure, once in place, has a tendency to precede the politics — and to create, in its functioning, the conditions that make the politics more, rather than less, possible.
That is what cyber diplomacy looks like when it is working quietly, in the language of telecommunications rather than statecraft, at the infrastructure level where routes are either open or they are not.
Armenia · Azerbaijan · AzerTelecom · Telecom Armenia · South Caucasus · Cyber Diplomacy · Digital Infrastructure · Conflict Resolution · Nagorno-Karabakh · Confidence Building · Vladimir Tsakanyan
Leave a comment